Deep learning in healthcare with improved architecture and representation learning
Deep learning has been crucial in recent times as many software and applications are using deep learning algorithms for tasks involving as image classification and Electrocardiogram (ECG) classification. Numerous deep learning architectures have also been introduced and studied over the years to pro...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Final Year Project |
| Language: | English |
| Published: |
Nanyang Technological University
2020
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/144844 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Nanyang Technological University |
| Language: | English |
| Summary: | Deep learning has been crucial in recent times as many software and applications are using deep learning algorithms for tasks involving as image classification and Electrocardiogram (ECG) classification. Numerous deep learning architectures have also been introduced and studied over the years to provide the high performing model architecture to be trained and deploy for the applications. Among the deep learning architecture, residual network (ResNet) is one of the best performing architecture that is widely used in the industry. Thus, this paper will explore and potentially improve the residual network architecture. Moreover, representation learning will be done to visualize the decision boundary that can be drawn from the features extracted by the proposed model. |
|---|